[Bug c++/99331] New: -Wconversion false-positive in immidiate context

[Bug c++/99331] New: -Wconversion false-positive in immidiate context

[nok.raven at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

            Bug ID: 99331
           Summary: -Wconversion false-positive in immidiate context
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Keywords: diagnostic
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: nok.raven at gmail dot com
  Target Milestone: ---

template <int N> struct X {};
template <class T> X<sizeof(T)> foo();

g++ -Wconversion -Werror src.cpp

<source>:2:22: error: conversion from 'long unsigned int' to 'int' may change
value [-Werror=conversion]
    2 | template <class T> X<sizeof(T)> foo();
      |                      ^~~~~~~~~

https://godbolt.org/z/d16WhG

[Bug c++/99331] [8/9/10/11 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mpolacek at gcc dot gnu.org
     Ever confirmed|0                           |1
            Summary|-Wconversion false-positive |[8/9/10/11 Regression]
                   |in immidiate context        |-Wconversion false-positive
                   |                            |in immediate context
   Target Milestone|---                         |8.5
           Priority|P3                          |P2
   Last reconfirmed|                            |2021-03-01
             Status|UNCONFIRMED                 |NEW

--- Comment #1 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Confirmed.  Started in GCC 7.

[Bug c++/99331] [8/9/10/11 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |mpolacek at gcc dot gnu.org

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Started with

commit 16b61424dd309f61326f577a6deb8487c6c1f291
Author: Jason Merrill <jason@redhat.com>
Date:   Fri Oct 21 15:45:45 2016 -0400

    re PR c++/77656 (64-bit integral template parameter gets incorrectly sized
as 32-bits)

            PR c++/77656
            * pt.c (convert_template_argument): Call convert_nontype_argument
            on value-dependent but not type-dependent arguments.
            (convert_nontype_argument): Handle value-dependent arguments.
            (canonicalize_expr_argument): New.
            (deducible_expression, unify): Skip CONVERT_EXPR.
            * error.c (dump_template_argument): Likewise.
            * mangle.c (write_expression): Likewise.

    From-SVN: r241425

think I'll have a look at this myself, though.

[Bug c++/99331] [8/9/10/11 Regression] -Wconversion false-positive in immediate context

[nok.raven at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #3 from Nikita Kniazev <nok.raven at gmail dot com> ---
This one most likely has the same root problem:

template <int N> struct X {};
template <typename T>
struct foo { using t = X<sizeof(T)>; };

<source>:3:26: error: conversion from 'long unsigned int' to 'int' may change
value [-Werror=conversion]
    3 | struct foo { using t = X<sizeof(T)>; };
      |                          ^~~~~~~~~

[Bug c++/99331] [8/9/10/11 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #4 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
This is caused by this change:

@@ -7278,7 +7306,7 @@ convert_template_argument (tree parm,
          val = error_mark_node;
        }
    }
-      else if (!dependent_template_arg_p (orig_arg)
+      else if (!type_dependent_expression_p (orig_arg)
           && !uses_template_parms (t))
    /* We used to call digest_init here.  However, digest_init
       will report errors, which we don't want when complain

Here orig_arg is SIZEOF_EXPR<T>; dependent_template_arg_p (orig_arg) was true,
but  type_dependent_expression_p (orig_arg) is false so we warn in
convert_nontype_argument.

[Bug c++/99331] [8/9/10/11 Regression] -Wconversion false-positive in immediate context

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Marek Polacek <mpolacek@gcc.gnu.org>:

https://gcc.gnu.org/g:9efd72d28956eb79c7fca38e3c959733a3bb25bb

commit r11-7833-g9efd72d28956eb79c7fca38e3c959733a3bb25bb
Author: Marek Polacek <polacek@redhat.com>
Date:   Thu Mar 4 20:20:40 2021 -0500

    c++: -Wconversion vs value-dependent expressions [PR99331]

    This PR complains that we issue a -Wconversion warning in

      template <int N> struct X {};
      template <class T> X<sizeof(T)> foo();

    saying "conversion from 'long unsigned int' to 'int' may change value".
    While it's not technically wrong, I suspect -Wconversion warnings aren't
    all that useful for value-dependent expressions.  So this patch disables
    them.  This is a regression that started with r241425:

    @@ -7278,7 +7306,7 @@ convert_template_argument (tree parm,
              val = error_mark_node;
            }
        }
    -      else if (!dependent_template_arg_p (orig_arg)
    +      else if (!type_dependent_expression_p (orig_arg)
               && !uses_template_parms (t))
        /* We used to call digest_init here.  However, digest_init
           will report errors, which we don't want when complain

    Here orig_arg is SIZEOF_EXPR<T>; dependent_template_arg_p (orig_arg) was
    true, but type_dependent_expression_p (orig_arg) is false so we warn in
    convert_nontype_argument.

    gcc/cp/ChangeLog:

            PR c++/99331
            * call.c (build_converted_constant_expr_internal): Don't emit
            -Wconversion warnings.

    gcc/testsuite/ChangeLog:

            PR c++/99331
            * g++.dg/warn/Wconversion5.C: New test.

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[8/9/10/11 Regression]      |[8/9/10 Regression]
                   |-Wconversion false-positive |-Wconversion false-positive
                   |in immediate context        |in immediate context

--- Comment #6 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed on trunk thus far.

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[nok.raven at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #7 from Nikita Kniazev <nok.raven at gmail dot com> ---
The fix silenced the true warning (though it was saying 'may') in these:

template <char N> struct X {};
template <class T> X<sizeof(T)> foo();
int x = sizeof(foo<char[512]>());


template <char N> struct X {};
template <typename T>
struct foo { using t = X<sizeof(T)>; };
foo<char[512]> x;


Should I post this as a separate issue?

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #8 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Yeah, that's expected (but it's a bug!):
https://gcc.gnu.org/pipermail/gcc-patches/2021-March/566449.html

Opening a separate issue would be nice, thanks.

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

--- Comment #9 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Marek Polacek
<mpolacek@gcc.gnu.org>:

https://gcc.gnu.org/g:b281edd801018ed703ce7bb0d49fef59ff8cef8b

commit r10-9688-gb281edd801018ed703ce7bb0d49fef59ff8cef8b
Author: Marek Polacek <polacek@redhat.com>
Date:   Thu Mar 4 20:20:40 2021 -0500

    c++: -Wconversion vs value-dependent expressions [PR99331]

    This PR complains that we issue a -Wconversion warning in

      template <int N> struct X {};
      template <class T> X<sizeof(T)> foo();

    saying "conversion from 'long unsigned int' to 'int' may change value".
    While it's not technically wrong, I suspect -Wconversion warnings aren't
    all that useful for value-dependent expressions.  So this patch disables
    them.  This is a regression that started with r241425:

    @@ -7278,7 +7306,7 @@ convert_template_argument (tree parm,
              val = error_mark_node;
            }
        }
    -      else if (!dependent_template_arg_p (orig_arg)
    +      else if (!type_dependent_expression_p (orig_arg)
               && !uses_template_parms (t))
        /* We used to call digest_init here.  However, digest_init
           will report errors, which we don't want when complain

    Here orig_arg is SIZEOF_EXPR<T>; dependent_template_arg_p (orig_arg) was
    true, but type_dependent_expression_p (orig_arg) is false so we warn in
    convert_nontype_argument.

    gcc/cp/ChangeLog:

            PR c++/99331
            * call.c (build_converted_constant_expr_internal): Don't emit
            -Wconversion warnings.

    gcc/testsuite/ChangeLog:

            PR c++/99331
            * g++.dg/warn/Wconversion5.C: New test.

    (cherry picked from commit 9efd72d28956eb79c7fca38e3c959733a3bb25bb)

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #10 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed in GCC 10.4 too.

[Bug c++/99331] [8/9/10 Regression] -Wconversion false-positive in immediate context

[ppalka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99331

Patrick Palka <ppalka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jens.maurer at gmx dot net

--- Comment #11 from Patrick Palka <ppalka at gcc dot gnu.org> ---
*** Bug 80601 has been marked as a duplicate of this bug. ***